Agriculture Reference
In-Depth Information
citrus harvesting robot (IVIA, 2004). This robot is different from weeding or scout-
ing robots as it has an on-board manipulator to identify and harvest citrus fruits.
Similar research efforts to develop citrus harvesting robots were conducted at the
University of Florida by Hannan et al. (2004).
Robotic harvesters for specialty crops such as cherry tomatoes (Kondo et al.,
1996), cucumbers (van Henten et al., 2002), mushrooms (Reed et al., 2001), cher-
ries (Tanigaki et al., 2008), and other fruits (Kondo et al., 1995) have also been
developed. Although autonomous robotic manipulators are commercially available
for milking and horticultural applications, mobile field robots are still not commer-
cially available. The most sophisticated tractors available today feature automation
of numerous machine functions but require an operator to closely monitor the tasks
being performed.
The successful implementation of autonomous systems by researchers and equip-
ment manufacturers is testimony to the fact that autonomous agricultural vehicles
can work in real-world applications, and the field of agriculture is evolving in to
a high-technology work environment. Although autonomous, these first-generation
systems require close supervision by human operators and further improvements are
needed to transform them into intelligent autonomous machines.
5.10 NEXT GENERATION OF AUTONOMOUS FIELD MACHINERY
Efforts to date have focused on removing much of the control from the human opera-
tor. Commercially successful devices in the marketplace actually extend the produc-
tivity of their human minders, making it possible to farm even more acres per unit of
labor. Along the development trail, many of the technologies necessary for automat-
ing agricultural machine functionality—precision metering and placement of pro-
duction inputs, and the harvesting and field processing of agricultural products—are
being perfected. What remains between today's machinery market offerings, and
fully autonomous field machinery, is the ability for multiple machines to interact
in a coordinated fashion to accomplish a variety of field production activities. The
following discussion, much of which is based on the work of Pitla (2012), provides
a roadmap of machine functionality to move field machinery toward unsupervised
operation.
5.10.1 I
NDIVIDUAL
R
OBOT
C
ONTROL
A
RCHITECTURES
Most of the initial work done on control architectures of mobile robots was car-
ried out in the aerospace and artificial intelligence research laboratories to accom-
plish military missions and space explorations. Unlike industrial robots, where the
environment is controlled and structured, the work environment of Ag-Robots is
relatively unstructured, unpredictable, and dynamic. An intelligent, robust, and
fault-tolerant control architecture is essential to ensure the safe and desired operation
of the Ag-Robot. A behavior-based (BB) control approach provides an autonomous
mobile robot the intelligence to handle complex world problems using simple behav-
iors. Complex behaviors of a robot emerge from simple behaviors (Brooks, 1986),
behavior being defined as response to a stimulus (Arkin, 1989). BB control structure
